[JaxPM] udp to tcp proxy

[Nate Campi]

On the jacksonville-pm-list; Jax.PM'er Nate Campi <nate at campin.net> wrote -

I have a syslog proxy I wrote that takes syslog packets in via UDP and
writes them to a SSL tunnel (via stunnel) to a central syslog collection
box.

Sometimes two or more of my writes to the TCP socket run together (four 
in this example):

Apr 9 - 2 times(s): 192.168.51.137 netstat: [USER.NOTICE]: gethostbyaddr: bigip2.bigip.hotbot.com. != 192.168.48.3 <13>Apr 9 07:33:35 192.168.51.137 netstat[20149]: [USER.NOTICE]: gethostbyaddr: music-ap.hotwired.com. != 192.168.49.60 <13>Apr 9 07:33:35 192.168.51.137 netstat[20149]: [USER.NOTICE]: gethostbyaddr: hbpdclaw01.hotwired.com. != 209.202.223.119 <146>Apr 9 07:33:35 capital mon[4493]: [LOCAL2.CRIT]: failure for monitoring_boxes ping 1018362815 nameb.hotwired.com

I add the facility/severity pair to the messages so the info lives in
the log archives, so this output may look a little weird to you.

Anyways, this is on a solaris 2.6 (sparc) box. I have three such
proxies, the other two on linux/x86. All are perl 5.6.1 and only the
solaris box exibits this behavior.

Has anyone else seen anything like this? I don't think it's the code
since it only happens on the solaris box. It could be something wrong
with stunnel, or another factor, but I wanted to know if this has been
encountered before by someone on the list.
-- 
"Error reading FAT record. Try the SKINNY one? (Y/N)"  -Anon.  


Jax.PM Moderator's Note:
This message was posted to the Jacksonville Perl Monger's Group listserv.
The group manager can be reached at -- owner-jacksonville-pm-list at pm.org
to whom send all praises, complaints, or comments...